The ovaries from slaughtered Peppin Merino ewes were examined at Trangie for a 12 month period commencing in April 1954. A seasonal rhythm was demonstrated for the percentage of ewes ovulating and for the percentage of multiple ovulations. The seasonal peak was in March–April, with a trough in November–December. Data were compared from a series of autumn and spring matings over the years 1953–1959 using genetically comparable flocks of similar age distribution, mated by identical methods. The overall incidence of oestrus was significantly higher in the first 3 weeks of mating (artificial insemination) in the autumn. The patterns of incidence of oestrus also differed, oestrus being much higher at the commencement of mating in the autumn. Differences between years in the incidence of oestrus were more pronounced in the spring, low incidence being correlated with poor nutrition. In only one mating — spring of 1954, a drought year — was the incidence of oestrus low initially, with a sharp rise commencing at 14 days. The conception rate for autumn-mated ewes was 10 percent. higher than for spring-mated ewes. Little difference was observed in the number and motility of sperm used at the seasonal matings. Lambing percentages strongly favoured the autumn mating: 15 percent. more wet ewes, 25 per cent. more twins mothered, 37 per cent. more lambs mothered, and 33 per cent. more lambs weaned. There was little difference between seasons in ewe deaths prior to and during lambing, or in deaths of lambs between mothering and weaning. The relative importance of the factors causing the 37 per cent. more lambs mothered in the autumn mating was as follows: incidence of oestrus, 11 per cent.; pattern of oestrus, 5 per cent.; incidence of multiple ovulation, 57 per cent.; conception rate, 27 per cent. The importance of the conception rate could be largely an associated effect of the higher autumn incidence of multiple ovulation. The weaning weight of lambs tended to be slightly higher for autumn matings, although there was great variation between years, both in the size and in the sign of differences. These differences had almost disappeared by 17 months of age. The choice of a mating time in central western New South Wales is discussed relative to these findings.